Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
In-Reply-To: <Y8FwGTWeYtX0j8MX@ravnborg.org>
References: <20230106030108.2542081-1-swboyd@chromium.org> <Y7nV+aeFiq5aD0xU@ravnborg.org>
 <CAE-0n50QOv_+j1Pe19xKj4Cx2Y5_Ak5Kt68UBJuZt10D-jQ44g@mail.gmail.com> <Y8FwGTWeYtX0j8MX@ravnborg.org>
From:   Stephen Boyd <swboyd@chromium.org>
User-Agent: alot/0.10
Date:   Fri, 13 Jan 2023 14:49:43 -0600
Message-ID: <CAE-0n51XmyzSeKuGNn2BsJew3fK_cEHNoDpVVRqStNAF+Zzp1A@mail.gmail.com>
Subject: Re: [PATCH] drm/panel: boe-tv101wum-nl6: Ensure DSI writes succeed
 during disable
To:     Sam Ravnborg <sam@ravnborg.org>
Cc:     Thierry Reding <thierry.reding@gmail.com>,
        linux-kernel@vger.kernel.org, patches@lists.linux.dev,
        dri-devel@lists.freedesktop.org,
        linux-arm-kernel@lists.infradead.org,
        linux-mediatek@lists.infradead.org,
        yangcong <yangcong5@huaqin.corp-partner.google.com>,
        Douglas Anderson <dianders@chromium.org>,
        Jitao Shi <jitao.shi@mediatek.com>,
        Rob Clark <robdclark@chromium.org>,
        Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Quoting Sam Ravnborg (2023-01-13 06:52:09)
> Hi Stephen,
> On Tue, Jan 10, 2023 at 11:29:41AM -0800, Stephen Boyd wrote:
> > Quoting Sam Ravnborg (2023-01-07 12:28:41)
> >
> > >
> > > For this case we could ask ourself if the display needs to enter sleep
> > > mode right before we disable the regulator. But if the regulator is
> > > fixed, so the disable has no effect, this seems OK.
> >
> > What do you mean by fixed?
> What I tried to say here is if we have a fixed regulator - or in others
> words a supply voltage we cannot turn off, then entering sleep mode is
> important to reduce power consumption.
> But any sane design where power consumption is a concern will have the
> possibility to turn off the power anyway.

Ok got it!

>
> >
> > >
> > > Please fix the unprepare to not jump out early, on top of or together
> > > with the other fix.
> >
> > After this patch the unprepare only bails out early if the bool
> > 'prepared' flag isn't set.
> OK, then everything is fine.
>

Doug pointed out that enable isn't symmetric because it doesn't do the
DSI writes. I've updated the patch and I'll send a v2.